A cost-sensitive extension of AdaBoost with markov random field priors for automated segmentation of breast tumors in ultrasonic images.
Atsushi TakemuraAkinobu ShimizuKazuhiko HamamotoPublished in: Int. J. Comput. Assist. Radiol. Surg. (2010)
Keyphrases
- markov random field
- cost sensitive
- breast tumors
- ultrasonic images
- multi class
- maximum a posteriori
- spectrum analysis
- automated detection
- pairwise
- ultrasound images
- parameter estimation
- higher order
- numerical data
- graph cuts
- random fields
- image restoration
- image segmentation
- digital mammograms
- active learning
- energy function
- conditional random fields
- object detection
- bayesian framework
- frequency domain
- breast cancer
- face detection
- autoregressive
- automated analysis
- learning algorithm
- naive bayes
- support vector machine
- segmentation method
- support vector
- training data
- multi label
- prior knowledge
- multiresolution
- adaptive filter
- image processing
- feature selection